Neuroscientists rely on animal models for the investigation of basic neural processes. Within this broad field, neuroethology has focused on the analysis of natural behaviors, often those associated with sensory processing of critical signals. Our understanding of the neuroanatomy, physiology, and development of human sensory systems comes largely from study of vertebrate model system. Glymnotiform electric fish have become leading models for investigation of parallel processing and organization among neural networks in vertebrate sensory and motor systems. Neuroscientists have traced the electrosensory pathway from the sensory periphery, through the brain, and back to the motor output. Gymnotiform electric fish use the same neural networks for three different functions: electrolocation, communication, and jamming avoidance. The gymnotiform electrosensory system is therefore an excellent model for exploring the flexibility of parallel processing networks. Behavioral studies of these three sensory functions form the base for investigation of the properties of the brain networks involved. This study is designed to reveal the behavioral mechanism(s) of electric signal discrimination involved in social communication. Further investigation of underlying networks and neural mechanisms can follow once we know the behavioral algorithms that electric fish use to discriminate among signal waveforms. Gymnotiform fish appear to control their motor output (electric signals) to enhance their sensory processing of signals from other fish. Combined signals sum to overcome inertia of the peripheral receptor system but must be separated once again by the central networks. The study proposed here follows a rational progression. 1. A calibrated set of natural signals will be digitized from the fish in an outdoor breeding colony for use as stimuli in subsequent experiments. 2. The next step will be to calibrate the sensory thresholds of the study species by obtaining two types of behavioral tuning curves: sensitivity to pure signals and the sensitivity to combined signals of receiver and signaler. 3. Signal motor patterns will be recorded in experiments that elicit natural discrimination behavior. Motor behavior in this context will be subjected to detailed analysis to document precisely when and how the receiver overlaps its signal with that of the signaler (i.e., combines signals). 4. The final experiments in this proposal will reveal the salient features used for signal discrimination and the motor behaviors required for accurate discrimination of signals.